Most of the Coalition spending and savings measures announced yesterday are pretty self-explanatory, but there are some relatively obscure changes included in the mix: savings from the contingency reserve, and a higher efficiency dividend on the public service.

The contingency reserve is not a fund of money.  It is an estimating device.  It was introduced to make the forward estimates more accurate.  It is explained in the hitherto obscure Attachment B to Budget Statement 6 of Budget Paper 1.

Contrary to some reports, the contingency reserve does not include an allowance for government department overspends — in fact it is the reverse, the reserve includes an allowance for the observed tendency of departments not to be able to spend all of the program funds they receive. It also includes some sensitive estimates, such as national security, that cannot be published  explicitly, and an allowance for state/territory agreements that are yet to be renegotiated.

The main component of the reserve is a correction for “conservative bias” in the estimates, the tendency for expenses to be greater in the forward years.  This bias is largely an artefact of the way the estimates are constructed — they include several programs that technically are due to lapse (for example, because a fixed-term funding agreement expires). We know that in reality governments do not let all lapsing programs  expire — they renew many of them.  The further out in the forward estimates you go, the larger the contingency reserve gets, because the number of technically lapsing programs is cumulative each year.  It is in these later years from 2012-13 that the Coalition makes its savings.

It is a bit like the estimate included in the budget for natural disasters (which also appears in the same function in the budget papers).  The officials preparing the estimates know there are likely to be some natural disasters, but have no way of knowing what these will be or how much the relief work will cost, so they include an estimate in the budget based on an observation of the past trends. The contingency reserve is likewise largely based on empirical observation of what has actually happened to the estimates in the past.

It is therefore not impossible for the Coalition to make savings from the contingency reserve — just unlikely.  It would have to take a much more aggressive approach to terminating lapsing programs than has any other government, Coalition or Labor, since the introduction of the forward estimates in the 1980s, and take a hard line on any other estimates changes to programs.

To be fair to the Coalition, the details of components of the reserve are not published in the budget papers — and so there may be scope for savings if it is prepared to break with the practices of past governments.  Only the finance department will have all the details.  If the Coalition wins, finance will have that list ready in its incoming government briefing folder.

The increase in the efficiency dividend from 1.25%-2% is doable.

The Rudd government raised it to 3.25% as a temporary measure in its first budget.

Two per cent is not even particularly radical.  If you want radical, you go back to the National Commission of Audit report in 1996 that recommended “minimum across the board efficiency targets of at least 10% … for the administration of all Commonwealth programs … in addition to the ongoing annual efficiency dividend” (recommendation 5.11).  That was one recommendation never adopted, even in the savings rounds of the early Costello budgets.

The efficiency dividend was introduced to share public sector productivity gains between public service agencies and the Australian taxpayers. Roughly half of the savings from greater productivity were to be retained by agencies, the other half returned to help improve the budget bottom line.

That was back when any negotiated public service pay rises were separately added on to agency budgets.  The original logic behind the efficiency dividend subsequently broke down when policy changed so that pay rises had to be paid for out of productivity gains; it is now in effect an across the board savings measure on the departmental budgets (that is, the budgets for salaries and administration) of public service agencies.

In net terms it has not even reduced public service budgets — large and powerful agencies have been able to recoup all of the departmental funding lost from dividend (and more) through their new policy bids.  Small agencies without clout and influence in the budget process have therefore been hit hardest by the dividend.

The recent public service review,  Ahead of the Game — Blueprint for the Reform of Australian Government Administration recommended a review of the efficiency dividend, to “address concerns expressed about unintended impacts of the efficiency dividend and its interaction with additional efficiency requirements in areas such as IT, central purchasing and property”.

The problem with the efficiency dividend is that it is not strategic — it hits all agencies the same, regardless of their circumstances. The alternative is a more targeted program of savings, based on analysis of where the greatest scope for savings lies.  That’s a more logical approach that would lead to better program outcomes, but perhaps not better political outcomes — government would have to specify where the cuts were being made.  That is often a challenge for ministers.

So the efficiency dividend saving is achievable and a real saving.

It is though not a great way to bring about a more efficient public service.  There is plenty of scope for efficiency: but an alternative way to achieve it would be through rigorous assessment of each agency’s performance and resourcing, tougher scrutiny of new policy proposals, and a reduction in the number of new programs and new regulation (all of which require people to administer them).